cutgeo for 1D periodic system

[sunxl]

Dear all,

I am calculating the GW and BSE for a one-dimensional helical chain with the z-axis as the periodic direction. How should I set up CUTGeo? Should it be set as slab XY or ws Z? May I ask what ws Z setting represents? Are there any differences between these two?

Best,
sunxl

[Daniele Varsano]

Dear Sunxl,

for 1D system periodic along z you have three options:
1) Box XY setting the cutoff size on the non-periodic directions (set it slightly smaller than your cell size)
2) Cylinder Z setting the cylinder radius
3) ws Z (wigner seitz cut), need to set an energy cutoff on the coulomb component to be corrected (default 0.7 a.u)

I discourage the use the cylinder as it can be unstable. My suggestion is to use "WS Z" but this is viable only if your cell is orthorhombic. Please note that for this option convergence should be checked against k point sampling, but being 1D usually one can afford a large sampling.

Best,
Daniele

[sunxl]

Dear Daniele，
Thank you for your reply.

For the first case, if I set box XY, I need to specify the size of the box in the x and y directions respectively. If my crystal lattice is hexagonal and the angle between ab is 120°, where a is parallel to x, and assuming that my a and b are both 30 au, then

1. Is it appropriate to set the box in the x direction to 29 or 28?
2. Since y and b are not in the same direction, how should I set the size in the y direction to be appropriate?

Best,
sunxl

[Daniele Varsano]

Dear sunxl,

the box cut value are appropriate, anyway as you noticed the potential cut geometry is not anymore consistent with your ab cell.
Possibly this is not a major issue if you have enough vacuum, but this should be checked. Alternatively, I suggest you resort to an orthogonal ab cell even if you would have more vacuum for the same image distance.

Best,

Daniele